Shae McFadden

PhD Student @KingsCollegeLon, focusing on machine learning applications for cybersecurity.

[Augustine Mavor-Parker @MavorParker]

Training a model to generate RL tasks not too hard, not too easy costs many solver runs per task.

PROPEL predicts difficulty via a probe on its activations instead, amortizing cost and speeding up generator optimization.

New open-ended RL research from @Vmax + @GoodfireAI.

[The Alan Turing Institute @turinginst]

💻 Malware is evolving. Can our defences keep up?

In collaboration with @ucl & @KingsCollegeLon, we’ve demonstrated how Deep Reinforcement Learning (DRL) agents can improve malware detection beyond conventional supervised learning approaches.

Learn more: https://t.co/5TnrpQG3UO

[Shae McFadden @Shae_McFadden]

I am excited to share that our paper "DRMD: Deep Reinforcement Learning for Malware Detection under Concept Drift" has been accepted to AAAI 2026.

https://t.co/Qg4ETRVR1Z

Co-Authors @0x6d66, @mario_donghia, @hkscy, @mavroudisv , Nicola Paoletti, and @fbpierazzi .

#AAAI2026

[Shae McFadden @Shae_McFadden]

This work was performed in collaboration with Marcello Maugeri, @hkscy, @mavroudisv, @fbpierazzi. #ieeesp

[Shae McFadden @Shae_McFadden]

I am happy to announce that I will be attending IEEE S&P to present our paper “Wendigo: Deep Reinforcement Learning for Denial-of-Service Query Discovery in GraphQL” at the DLSP workshop.